1. Introduction
Indirect reports, besides being multidirectional and multifunctional, are hybrid constructs, to employ Mikhail Bakhtin’s terminology. Bakhtin defines a hybrid construction as ‘an utterance that belongs, by its grammatical (syntactic) and compositional markers, to a single speaker, but that actually contains mixed within it two utterances, two speech manners, two styles, two “languages”, two semantic and axiological belief systems’ (Bakhtin & Holquist, Reference Bakhtin and Holquist1981, p. 304). The concept of the hybrid structure manifests that every utterance can act as an indirect report since, if one takes Bakhtin’s notion of dialogue into consideration, the interaction between individuals is under the influence of sociopragmatic idiosyncrasies attached to contextual factors, shaped and guided by the presence of others. Although indirect reports are not necessarily hybrid constructions, it can be argued that the use of language is fundamentally polyphonic, an idea corroborated by Authier-Revuz (Reference Authier-Revuz1984). In this regard, Nølke (Reference Nølke2006, p. 137) refers to the duality of voices that exists ‘not only in parole but in langue’.
For Bakhtin, polyphony is not reduced to the grammatical and stylistic problems of mixing two sentences (Banfield, Reference Banfield1982) but is a discourse phenomenon where utterances, speech manners, styles, languages, semantic and axiological belief systems are intermingled, reminiscent of hybrid constructions. In line with the belief systems, Bakhtin differentiates three types of double-voiced discourse (DVD). Regarding the first type, there is cooperation between the author’s (reporter/narrator) discourse and that of the other person (the original speaker). In the second type, the author’s discourse is opposed to that of the other person (Zbikowski, Reference Zbikowski2002). In this case, Bakhtin (Reference Bakhtin1984, p. 193) points out that ‘the second voice, once having made its home in the other’s discourse, clashes hostilely with its primordial host and forces him to serve directly opposing aims. Discourse becomes an arena of battle between the two voices’ (emphasis added). In this second type of discourse, the perspective of the reporter is in contrast with that of the original speaker. There is still a third type of DVD where ‘the other person’s discourse remains outside the purview of the author’s, which nonetheless becomes influenced by that of the other person’ (Zbikowski, Reference Zbikowski2002, p. 225). The author only indirectly reacts to the discourse of the other person (hidden polemic).
More often than not, when speaking of ‘two voices’ and ‘two languages’ existing in discourse, it is typically referring to interactants with the same native language, thereby sharing the same linguistic code, with variations between voices being attributed to differences in beliefs, language use, cognitive flexibility, and so forth. It is apparent that the possibility of space between multiple voices only grows when speakers are not speaking in their native language. Non-native capacity in a language adds opportunities for a separation between voices. According to the Interface Hypothesis (Sorace, Reference Sorace2011; Sorace & Filiaci, Reference Sorace and Filiaci2006), certain interfaces between aspects of language in L2 such as syntax and discourse may be mastered by these speakers to the same degree of proficiency as native speakers. However, other interfaces such as those related to syntax with other cognitive systems are more difficult to master, even for highly accomplished L2 speakers.
What justifies the investigation of the reported speech by L2 learners is the fact that the learners’ ability to produce and comprehend various utterances in interaction manifests one aspect of the learner’s pragmatic competence. Moreover, through the analysis of indirect reports among the learners of L2, it is possible to discover how the illocutionary force of joint activities (e.g. indirect reports) is co-constructed among them (cf. Taguchi, Reference Taguchi2019). In addition, to highlight the importance of interaction in the contexts where English is an additional language, Murray (Reference Murray2012, p. 318) mentions that ‘[b]oth in the case of nativised varieties and the global use of English as a lingua franca, we are of course not dealing with one monolithic, hegemonic English voice but with a great diversity of different voices’. To create reported speech, syntactic rephrasing is often carried out on the part of the speaker that may not be problematic to an L2 speaker with a high level of competency; however, the processes Günthner referred to as ‘decontextualising’ and ‘recontextualising’ speech necessarily go beyond purely linguistic mechanisms, calling upon their interface with other cognitive systems. For example, the generation and processing of presuppositions in a sentence, which rely on cognitive systems as well as linguistic ones, have been shown to be different in L1 and L2 adults (Cho, Reference Cho2017; Evans et al., Reference Evans, Bergqvist and San Roque2018; Feng, Reference Feng2021). In this regard, Wong (Reference Wong2010, p. 2932) argues that
[…] what language is about goes beyond lexicon and grammar to include non-formal features like conversational routines, frequency of use of certain expressions, the avoidance of certain ways of speaking, pragmemes, etc., which can only be satisfactorily explained with reference to culture.
With regard to Bakhtin’s concept of DVD, this article illustrates that L2 indirect reports do not always abide by a fair paraphrase (i.e., the retelling of what the original speaker said) of the original speaker, for the learners can infuse their subjective manipulation from several sources into the original speaker’s discourse. In this sense, this article sheds light and elaborates on the claim that ‘the interests and the point of view of the reporter cannot prevail over those of the original speaker’ (Capone, Reference Capone2010, p. 381). Against this backdrop, this article makes an attempt to answer the following questions:
-
1. Do Iranian EFL learners change or maintain the syntactic/semantic/pragmatic features of the original sentences when indirectly reporting others in the classroom?
-
2. What is the nature of the possible syntactic/semantic/pragmatic alterations made by Iranian EFL learners when reporting others?
This study is done in an Iranian L2 setting. This context is important for the research on reported speech because reported speech is a common linguistic structure among languages. Moreover, culture-specific investigations have three main implications for a theory of reported speech. As implied by Kecskes (Reference Kecskes2014), the first is relevant to the significance of linguistic pragmatics for intercultural pragmatics. The concept of reported speech is linked with a theory of pragmatics. In accord with Güldemann and Roncador (Reference Güldemann, von Roncador, Güldemann and von Roncador2002, p. vii), ‘scholars have increasingly shifted their interest towards the question of how grammatical, semantic, and pragmatic problems associated with functional and formal properties of reported discourse have repercussions in other linguistic domains of language’. The second implication manifests the need to understand how linguistic structures are produced and comprehended collaboratively in intercultural encounters. Finally, the third implication is the need to analyse the common themes that may exist among languages when the grammatical system is utilised for the construction of the reported speech.
The outline of the study is as follows. In Section 2, direct and indirect reports are discussed in regard to their syntactic and semantic characteristics as well as the concept of polyphony and its relation to indirect reports. In Section 3, data and methodology are defined. In Section 4, findings are presented and discussed, focusing on the different types of DVD. Finally, in Section 5, concluding remarks are presented.
2. Background
2.1. Direct versus indirect reports
Research on direct and indirect reports has contributed significantly to pragmatics and discourse analysis (Earis & Cormier, Reference Earis and Cormier2013) by enriching the discussion on the semantics/pragmatics interface and by its connection with Bakhtin’s (Reference Bakhtin1984) concept of ‘polyphony’ (dialogic nature of discourse) and Wittgenstein’s notion of language games (reminiscent of social practice). Likewise, Nodoushan (Reference Nodoushan2015, p. 107) argues that considering ‘indirect reports on purely semantic grounds is doomed, and that a true interpretation of the truth-conditional nature of indirect reports will have to be based on an interplay between semantics and pragmatics’. The practice of indirect reports is linked with that of direct reports. Coulmas (Reference Coulmas1986) differentiates between two types of quotations, namely direct (oratio recta) and indirect (oratio obliqua). Direct quotations or reports are defined as using the exact syntactic structures of a speaker, woven into one’s speech. The direct reports are quotes from either written or spoken discourse. Individuals resort to direct reports as a powerful tool to ‘spice up’ their speech, appeal to authority, or justify a thought or idea. As stated by Ebert (Reference Ebert and Coulmas1986, p. 145), ‘[f]urther, thoughts, intentions and other cognitive acts are also represented in the form of direct speech’. Since the formation of direct reports needs higher cognitive processes (Kecskes, Reference Kecskes, Capone, Kiefer and Piparo2016), these constructs are not preferred in interactions (in comparison to indirect reports that give more structural freedom to speakers).
However, direct quotes are a more reliable and authoritative tool ‘because of their capacity to instantiate directly the expressive character of language’ (Lucy, Reference Lucy and Lucy1993, p. 19), which convey a more powerful impression on the interlocutors – assuming that the reporter is faithful to the original speaker’s utterances. Capone (Reference Capone2016, p. 54), likewise, advocates that ‘[i]ndirect reports seem to allow the speaker (and prepare the hearer for) a lower degree of accuracy than direct reports’. That said, although direct reports are easily recognisable in written interaction (they are always accompanied with quotation marks), they are not readily identifiable in authentic communication (Earis & Cormier, Reference Earis and Cormier2013). Li (Reference Li and Coulmas1986) has also discussed the structural differences between direct and indirect reports focusing on syntactic and semantic elements. In line with syntactic differences, direct (1) and indirect reports (2) are distinct in the following terms:
[F]irst, the pronouns in (1) and (2) are different; second, the tenses in (1) and (2) are different; third, (2) but not (1) may have the complementizer ‘that’. Finally,… the immediate constituent grouping in terms of phonological pause and intonation pattern is different between (1) and (2). (Li, Reference Li and Coulmas1986, p. 29)
Comparing direct and indirect reports, Capone (Reference Capone2016, p. 53) mentions that ‘they are mainly a matter of degree, although there are non-negligible syntactic differences, as direct reports admit interjections, while indirect reports, allegedly, do not’. Capone further argues that ‘in direct reports, especially those of the strict type, opacity is a result of interpreting the report verbatim; opacity is a pragmatic phenomenon in indirect reports’ (p. 53).
The indirect report is a tool to make conversational implicatures (by differentiating what is said from what is implied). Indirect reports are in accord with the maxim of quantity (Grice, Reference Grice, Cole and Morgan1975) by limiting the speaker/reporter to provide as much information as needed and no more. The maxim of quality urges the reporter to make a contribution that is truthful (this is violated in the case of untrue statements). The maxim of relation necessitates that the aim of the indirect report should be related to the aim of the conversation. In regard to the maxim of manner, the reporter should clearly distinguish his/her voice from that of the original speaker. Indirect reports can be conceptualised structurally by using Goffman’s theory of footing. When the speaker reports a sentence from his/her past encounters with others, the speaker tries to mention what is said in the past (a type of animator in Goffman’s terminology) and at the same time manifests himself/herself as the principal, that is, a person who accepts the responsibility for what he/she has uttered. Nonetheless, in indirect reports, different utterers (or enunciators) can be responsible for the illocutionary force of the utterance.
2.2. Polyphony: changes and samesaying
Bakhtin’s notion of polyphony inspired the theory of ‘enunciative polyphony’ proposed by Ducrot (Reference Ducrot1984). Polyphony at the utterance level refers to the plurality of voices that takes into account the speaking subject (sujet parlant), the locutor (locuteur) and the enunciator (énonciateur). Johansson and Suomela-Salmi (Reference Johansson, Suomela-Salmi, Zienkowski, Ostman and Verschueren2011, p. 89) define the terms:
[T]he first one is the empirical being in the world producing the utterance, the second is present in the discourse uttering the act and the third is the ‘voice’ that has responsibility over the opinion, the attitude etc. of the utterance.
In line with the concept of ‘polyphony’ and the Wittgensteinian notion of the language game, Capone (Reference Capone2016, p. 80) suggests that ‘polyphony is a language gameFootnote 1 that is embedded in the practice of indirect reporting. The game also consists in the way clues and cues are utilised to separate the voices of the participants’. In concomitant with the characteristics of societal pragmatics (Dascal, Reference Dascal2003; Mey, Reference Mey2001), Capone (Reference Capone2016) manifests the connection of voices and the role of the hearer that helps him/her to distinguish the voices in a polyphonic structure such as indirect reports. As stated by Capone (Reference Capone, Capone and Feit2013, p. 159), ‘[o]ne of the problems we encounter in the description of indirect reports is, in fact, that an apparently single utterance contains different voices/points of view’ That said, it is quite possible that the reporter’s perspective deviates from that of the original speaker, for example, in the case of ‘strategic ambiguity’ pointed out by Fløttum (Reference Fløttum2010). Günthner (Reference Günthner1999, p. 686) argues that ‘[i]n reporting past utterances, the speaker ‘decontextualises’ speech from its original co- and context and ‘recontextualises’ it in a new conversational surrounding’. These changes also contain functional intentions and communicative aims which can move away from the original speaker’s perspective. In this regard, Nodoushan (Reference Nodoushan2017, p. 80) states that ‘[h]uman agents who produce linguistic utterances have certain intentions (in so doing), which are mutually manifest for both themselves and their interlocutors’ Capone (Reference Capone2010, p. 382) also argues that ‘[i]n the case of indirect reports, we have got a speaker′, a hearer, and a speake’ (the original speaker); both speaker′ and speaker ‘take responsibility for the content of the that-clause in case all goes well, that is to say, in case the indirect reporting is felicitous’. It seems that Capone’s view of DVD assimilates more closely with Bakhtin’s definition of the first type of DVD.Footnote 2 The second type of DVD, on the other hand, indicates the possibility of a clash between the author’s and the original speaker’s perspective although it is difficult to find out who should take the responsibility for the report.
Closely linked to the notion of polyphony is the nature and scope of changes in indirect reports. The construction of indirect reports has always been engaged with changes (Evans et al., Reference Evans, Bergqvist and San Roque2018), be they syntactic, semantic or pragmatic. Indirect reports follow some rules that require paraphrases. For all intents and purposes, not every transformation is considered licit in indirect reports. The point of view put forward in indirect reports should be made explicit by providing sufficient syntactic clues and cues so that it would be possible for the hearer to distinguish between the voices intermingled in the report (Morady Moghaddam, Reference Morady Moghaddam2019; Morady Moghaddam & Capone, Reference Morady Moghaddam and Capone2020). However, the difficulty in doing this in a licit manner is much greater in L2 than in L1, as slight nuances can have unintended effects on semantics, syntax and pragmatics, all of which can in turn affect the nature of polyphony created. Liu and Zhong (Reference Liu and Zhong2016) demonstrated that even advanced L2 learners displayed non-native choices between synonyms in a forced choice experiment, in an examination of construal, the process by which a speaker chooses a given view and expression in a given context (p. 241).
Another perplexing feature related to the concept of polyphony is ‘samesaying’. Donald Davidson has been a key figure in the introduction of the term ‘samesaying’ (cf. Davidson, Reference Davidson1968). ‘Samesaying’ refers to the fact that ‘the report and the speech to be reported have some broad content in common’ (Capone, Reference Capone2016, p. 24). The concept of ‘samesaying’ shares similarities with the ‘out-of-court statement’ or ‘hearsay’. Thus, in order to perform a ‘hearsay’ function, one of the samesayers (the reporting speaker) should use the words of someone else (the original speaker). In order to completely acknowledge the concept of ‘samesaying’, one needs to figure out the ‘hearsay’ concept. The literature on reported speech has closely dealt with the issue of samesaying by attempting to answer the licit and illicit changes that the reporter is allowed to make (Itakura, Reference Itakura, Capone, Garcia-Carpintero and Falzone2018; Van der Wurff, Reference Van der Wurff, Güldemann and von Roncador2002). It is clear that to accurately carry out the practice of ‘samesaying’, precise comprehension of the first speaker’s message at multiple levels must exist. This may be particularly challenging to the reporting speaker who is speaking in an additional language.
3. Current study: setting and data
Speakers are engaged in strategicFootnote 3 choices that are made during the interaction (Candlin et al., Reference Candlin, Crichton and Moore2017). As stated by Wong (Reference Wong2010, p. 2942): ‘to attain a more holistic picture of language, one would need to see beyond form and even meaning to consider the cultural context of the language (or pragmemes) under study’. Few studies on L2 reported speech have been carried out and these few explorations are very recent (Diskin & Levey, Reference Diskin and Levey2019; Hatzidaki et al., Reference Hatzidaki, Santesteban and Duyck2018; LeBlanc, Reference LeBlanc2018), clearly showing the novelty of research on L2 reported speech. In one of these studies on L2 reported speech, Diskin and Levey (Reference Diskin and Levey2019) dealt with the quotative ‘be like’ in Dublin English. These researchers came to the realisation that while comparing L2 quotative system with L1 Dublin English, ‘not all L1 usage constraints are faithfully replicated by L2 speakers, indicating that the acquisition of the relevant constraints is incomplete, even in the case of advanced learners’ (p. 53). Hatzidaki et al. (Reference Hatzidaki, Santesteban and Duyck2018) analysed the reported speech production (e.g. Holly asked what Eric ate) related to two groups of proficient bilinguals whose native languages were Spanish and Dutch and their additional language was English. This study, as well as the study done by Diskin and Levey, focuses on the role of interference from L1 on the production of L2-reported speech. Hatzidaki et al. (Reference Hatzidaki, Santesteban and Duyck2018) suggested that, as it is related to reported speech production in the bilingual groups, ‘cross-linguistic syntactic differences by themselves suffice to induce language interference, and that the degree of similarity between the L1 and the L2 does not seem to modulate the magnitude of this effect’ (p. 489). This contributes to the notion that extra room for error in reported speech is more likely to exist for L2 learners than for native speakers.
LeBlanc (Reference LeBlanc2018), through a study that ‘contributes to an understanding of intertextuality and reported speech in teacher talk’ (p. 150), probed ‘projected ventriloquizing’, meaning in this case that the teacher adopts her students’ voices and projects them out into the classroom, creating and uttering hypothetical examples of her students’ imagined criticism. In this case, the teacher avoids one form of emotional labor through another. LeBlanc makes the case that although ventriloquism recontextualises the speaking context, its features, including prosody and deixis, remain the same as the original, or in this case the ‘imagined’ or ‘ventriloquised’ original. A very high degree of linguistic knowledge, including and surpassing syntax, is required to achieve this communicative form (Taguchi, Reference Taguchi2019).
LeBlanc’s study is of particular importance to the current one, as it is based on classroom interaction. This study builds on the literature of DVD in L2 learners and seeks particularly to better understand the nature of polyphony in an L2 by exploring the sociopragmatics of indirect reports. Of specific interest are instances where discord exists between the original voice and the reporting voice. Qualitative research is a highly appropriate tool for initial data exploration of this type, especially given the relatively small sample size and the allowance of creativity in data analysis on searching for underlying patterns.
In order to shed more light on the pragmatic competence of L2 learners, this study employed data gathered from observing 35 intermediate to advanced EFL learners participating in free English conversations in an institutional setting. The data were gathered before the COVID-19 pandemic. Consent was provided by the participants, who were aware that their conversations were being recorded, but who were not aware of the specific research goals and objectives. Each class lasted 90 minutes with the final 30 minutes being used for free discussion among the learners. During the free conversation sections (10 sessions, each section lasting 30 minutes), several challenging and controversial issues pertaining to Iran were discussed and critically analysed by the teacher (an experienced Iranian male teacher who had been teaching English as a foreign language for about 10 years) and the learners. The teacher and the learners (N = 35) discussed challenges, solutions and policies across specific domains (social, economic, political and educational issues). One of the researchers was present in all the classes and recorded the conversations (the teacher was not one of the researchers). There were at least two individuals in each talk (i.e., learner-learner or teacher-learner), which created an interactive atmosphere while sharing information in class. Overall, we observed and examined 300 minutes of free discussion over approximately 3 months. The exchanges between the learners (including the original speaker and the reporter) were in the form of ‘eyewitness-to-eyewitness’ (the original speaker was present during the interaction) and ‘eyewitness-to-outsider’ (the original speaker was absent from the interaction).
Classroom discussions were recorded using a mobile phone and were transcribed by the researchers. The reported speech was identified if (1) the learners used the complementiser ‘that’ in their sentences or (2) a direct reference was made to someone else’s speech. A summary is provided in Table 1:
Table 1. The details of this study’s data
In the next section, we deal with the analysis of indirect reports and how L2 learners change the syntactic and semantic features of utterances in line with the sociopragmatic features. In our analysis, we discuss several rudimentary issues essential to the production and comprehension of indirect reports. In this article, we focus on the linguistic diversity among Iranian EFL learners’ production and comprehension of indirect reports.
4. Results and discussion
The goal of the article was to characterise the types of reporting that are attested in EFL data. In this section, we analyse specific examples of polyphony in Iranian EFL learners’ indirect reports as manifested by cooperation between voices (the first type of DVD) or by a clash between voices (the second type of DVD). The criterion for this distinction is the degree of cooperation or conflict between the reporter’s interpretation and the original speaker’s perspective (both the speaker and the reporter are Iranian EFL learners) as identified by the approval or disapproval of the original speaker. However, this is not a dichotomous distinction, but rather a continuum that contains different degrees and different expressions (both overt and covert) of cooperation or confrontation.
4.1. The first type of DVD: the cooperation of two voices
Indirect reports are perfect examples of polyphonic utterances.Footnote 4 Indirect reports are generally marked by the complementiser that and a variety of reporting verbs such as ‘said’. Speakers can promote cooperation in their indirect reports by paraphrasing the original speech that does not go beyond the reported speaker’s pragmatic meaning. In this regard, the reporter alters the structure of the original sentence but, to resort to Kecskes’s (Reference Kecskes, Capone, Kiefer and Piparo2016) conceptualisation of indirect reports, the three key elements (illocutionary force of the original message, actual situational context, and the evaluative load of the original message) are kept unchanged as in the original. In cooperative polyphony, ‘the integrity of the original message [is preserved] by formally marking the boundaries between the main message and the embedded reported message’ (Kecskes, Reference Kecskes, Capone, Kiefer and Piparo2016, p. 10). In other words, polyphony happened at the utterance level whereby the original speaker does not disagree with the pragmatic force of the report, no matter what changes are made to the original words (hence, DVD is cooperative rather than opposing). If one, for example, says ‘John lied that he would help me’, the syntactic construction perfectly separates the main message from the reported message but there is definitely no ‘cooperation’ between the two voices. In this case, the change in syntax is not the main criteria to see whether the main message and the reported message are cooperative. The illocutionary force/meaning of the utterances should be cooperative as well.
When indirect reports are viewed in this way, the pragmatic force of the reporter’s speech is very close to that of the original speaker. The changes made to the original speech may be due to several reasons such as a reduction of the cognitive load on the part of the reporter (in this sense, reporters do not have to memorise the exact words of the original speaker), for rhetorical effect, due to misunderstandings resulted from imperfect hearing or lack of full attention, or a desire to simply portray the original speaker in a particular way. The exchanges between the original speaker (S) and the reporter (R) in Extracts 1–4 all embrace cooperative polyphony in that the reporter changes the syntax of the original speech cooperatively, while keeping intact the pragmatic force of the original speaker, who may be presentFootnote 5 (eyewitness-to-eyewitness) or absent (eyewitness-to-outsider) at the time of the report. The following exchanges between the original speaker and the reporter (all of which are immediately consecutive) are a selection of examples extracted from the data that reveal the first type of DVD among Iranian EFL learners. ‘S’ refers to the speaker and ‘R’ refers to the reporter, both being Iranian learners of English who are communicating about ‘companies in Iran’:
Extract 1.
S: If someone is the chief of a big company, he is not just responsible for the production.
R: Now you say, a person who thinks industrially thinks commercially as well.
In Extract 1, lexico-syntactic alteration is obvious in the report. The reporter changes ‘is not just responsible for the production’ to ‘thinks commercially as well’ in order to touch on the implicit information implied by the original speaker’s sentence (i.e., selling). The reporter indicates knowledge of ‘what is said’ as well as ‘what is implied’ during the interaction, and he shows this in his conversation with another person in the class. Indeed, the original speaker attempted to imply that the president of a company does not only deal with the production, but there are other aspects, such as selling the products, which are equally important. The reporter interprets ‘what is implied’ from the previous information the original speaker had provided, and hence mentions the original speaker’s intention in the report without creating any opposition between the original speech and the report. Likewise, in Extract 2, the reporter paraphrases the original speaker’s utterances without changing the reported speaker’s pragmatic meaning:
Extract 2.
S: Just recently, the increase in prices has started, especially the increase in the value of the building materials.
R: One issue that you mentioned in your speech is the increase in the construction fee.
In Extract 2, ‘the increase in the value of the building materials’ is reported by one of the learners as ‘the increase in the construction fee’, which is in line with the speaker’s intended meaning (i.e., asserting an increase in building costs conversationally implicates or pragmatically conveys an increase in construction fee). The reporter interprets that the ‘building materials’ are used for ‘construction’. For this reason, instead of repeating the exact words of the original speaker, the reporter adds his voice to the report by slightly changing (‘building materials’ to ‘construction’) the syntax of the original message. Although the syntax is tweaked, the underlying meaning in the original speech and the report is remained intact. Furthermore, the adaptation of verbal tense was observed among Iranian EFL learners:
Extract 3.
S: Creative and skillful managers are preferred over those who have just high degrees from universities without enough experience.
R: You mean that you prefer those experienced managers over those who have academic knowledge.
In Extract 3, the learner changes the passive verb ‘are preferred’ to an active verb ‘you prefer’. The reporter in (3) uses ‘you mean that …’ in the report, which indicates that the reporter is not sure about what is being reported. That said, ‘you mean that …’ is an indicator of the level of accuracy that can help the hearer to distinguish between a direct and an indirect report because it refers to the learners’ interpretation of the original speech.
The data of this study not only illustrated that EFL learners change the verbs in line with the original speaker’s perspective, but they can also add adjectives and other parts of speech to the report based on their analysis of the original utterance:
Extract 4.
S: My view is that the relationship between Europe and the US during Trump’s presidency has experienced serious fluctuations and recently there has been a commercial war between the two sides.
R: As you mentioned in your comment, Trump has created a big economic war with the Europeans.
In Extract 4, ‘commercial war’ is changed to ‘big economical war’ in the reported speech, through which the reporter adds ‘big’ to ‘commercial war’ and changes ‘commercial’ to ‘economic’. The reporter in Extract 4 has omitted the phrase ‘serious fluctuations’ and added the word ‘big’ to ‘economic war’. In this case, ‘big’ can be considered pragmatically equivalent to ‘serious fluctuations’.
The changes observed in Extracts 1–4 point out that the reporter can change the syntactic features without manipulating the illocutionary force of the original utterance (therefore, no opposition occurs between the voices). This is justified based on Sperber and Wilson’s (Reference Sperber and Wilson1986, p. vii) Relevance Theory, where the ‘principle of economy’ indicates that ‘[h]uman cognitive processes … are geared to achieving the greatest possible cognitive effect for the smallest possible processing effort’. Iranian L2 learners’ reported speech is not a hundred per cent similar to the original speech syntactically, as they are able and willing to change the words in line with the broad content of the original speech. What is clear from the extracts above is that L2 learners are manipulative when reporting the speech of others in the class, and they try to reduce the cognitive load of utterances by remembering the main points and transforming the original speaker’s utterance in accord with the original speaker’s perspective. In all of the preceding extracts, the original speaker’s intention and the illocutionary force of the utterance are given priority over subjective manipulations by the reporter beyond the original speaker’s perspective, yet the syntax of the report is changed for a myriad of reasons. By this account, considering the cooperation of the two voices, Iranian learners’ reported speech indicates that samesaying is a hundred per cent perfect, although there might be some syntactic differences between the original speech and the report.
4.2. The second type of DVD: an arena of battle between two voices
On the other hand, there are polyphonic utterances where the boundary between the two voices becomes blurred. In this sense, Fløttum (Reference Fløttum2010, p. 998) argues that ‘[t]he unclearness is particularly obvious when the sources of the different points of view that are introduced are not made explicit’. Fløttum mentions that ‘[t]he positive aspect of this kind of unclearness is that it allows multiple perspectives to coexist in the same text. It also facilitates the participation of voices which either represent obvious contrasts or clear-cut contradictions’ (Fløttum, Reference Fløttum2010, p. 998). A contradiction arises when we compare Fløttum’s idea of ‘unclear language use’ with Capone’s view on polyphony in indirect reports. The key passage is:
The problem, as I see it, is that reporting speech is a language-game of its own (viz. that of indirect reporting) that is severely constrained by the fact that it displays two voices, the reporter’s and the original speaker’s; moreover there is a tension between them, such that the interests and the point of view of the reporter cannot prevail over those of the original speaker. (Capone, Reference Capone2010, p. 381)
This study’s findings on indirect reports are more aligned with Fløttum’s perspective on unclear language use, and the data also support Borg’s (Reference Borg2012) view on free pragmatic enrichment, which highlights interpretations that are not linked with any lexico-syntactic features of the source message (therefore, there is a clash between the voices). These kinds of ‘shadow conversations’ (Irvine, Reference Irvine1996) were frequently observed among Iranian EFL learners that counterargue with perfect samesaying. In the following examples observed in this study’s data, the original speaker challenges or refutes the reporter’s interpretation, thereby attempting to amend the implicit arguments made by the reporter. The examples below clearly indicate that the broad content of the report and the original utterances can be mutually exclusive:
Extract 5.
S: In my view, authorities should do something to bring back the variety of stuff in the market, not just one particular company or brand.
R: You say there is a need for a more competitive market?
S: I want to say that we need good managers.
Extract 6.
S: This year, our city has had a better situation with regard to housing prices and renting conditions.
R: You said home prices are decreased in Tehran.
S: I did not say such a thing. But, overall, it is possible to say that there has not been a tangible increase in the housing price this year.
Extract 7.
S: I believe that they have done their best and we should give them more time to decide what works best for the educational system of the country.
R: In the previous section, you said that you disagree.
S: No, I said desirable and more desirable.
Extract 8.
S: Everybody deserves to be treated properly, and based on his/her qualifications. We need good regulations that allow everyone to have a decent income. This is good for our society.
R: One hint in the centre of your discussion was that if we are going to keep a manager who has many qualifications, we should pay a higher salary to him/her. Am I right?
S: No, no, I am not satisfied with this kind of interpretation. I am saying that authorities should settle the rules properly.
The indirect reports observed in Extracts 5–8 are all examples of structural alteration beyond the original speaker’s perspective (hence providing evidence against perfect samesaying). The original speaker in Extract 5 self-reports himself, using the phrase ‘I want to say…’ in order to correct the report and to indicate that the reporter has not properly conveyed the original speaker’s pragmatic meaning. Using ‘I want to say’, the speaker counterargues with the reporter and also reports his own previous utterance. In Extract 6, the original speaker vividly refutes the reporter’s interpretation, saying ‘I did not say such a thing’ in an attempt to oppose the inappropriate report. Another instance of self-report is observed in Extract 7 where the original speaker tries to correct the reported speech by adding further information about his intentions by uttering ‘I said…’. Finally, in Extract 8, the original speaker says ‘I am not satisfied with this kind of interpretation’, manifesting that the reporter is ‘not satisfied with this kind of interpretation’, hence opposing the pragmatic meaning of the report. In the instances discussed, the original speaker was present when the report was made, and hence it was possible to give feedback and express dissatisfaction with the reported speech.
It is useful to once again note that indirect reports are helpful in making conversational implicatures. Upon examining Extracts 5–8, it is apparent that clashing interpretations between the speakers are not due to difficulties in analysing semantic implicature, but instead because of the context-specific conversational implicatures. Hearing a speaker declares that ‘housing prices are better this year’, it seems fair to make the conversational implicature that ‘prices are lower’, as long as the speaker is not in the business of selling housing. However, conversational implicatures typically are not logically entailed, but only suggested, by the original statement, allowing the implicature to be what Grice calls ‘cancelable’ (Grice, Reference Grice, Cole and Morgan1975, p. 57). It is noteworthy that Extracts 5–8 all represent contradictions between two speakers following from different interpretations of conversational, as opposed to semantic or syntactic, implicatures. This supports the Interface Hypothesis (Sorace, Reference Sorace2011; Sorace & Filiaci, Reference Sorace and Filiaci2006) in that it suggests, even for advanced L2 learners, it may be cumbersome to make a connection between syntactic and external components such as pragmatics of L2 learning that requires the interface between language processing systems and other cognitive abilities (Stringer & Tommerdahl, Reference Stringer and Tommerdahl2015). In this case, higher-order cognitive abilities allow speakers to extrapolate new knowledge based on the given information.
The Extracts 5–8 acknowledge polyphony at the pragmatic level that contradicts the intended meaning of the original speaker’s utterance. For these examples, the data in this study refute the exactness of similarities between the quotation and paraphrase in indirect reports. The wordings that come after the complementiser ‘that’ may deviate from the original utterance if they are not in accord with the original speaker’s main intention. As it was mentioned above, Capone’s view on polyphony in indirect reports suggests that the interests and point of view of the reporter cannot prevail over those of the original speaker. That being the case, this study’s findings on L2 reported speech revealed that this premise is not correct since there were examples where the reported speech was different from the original speech because it lost its proximity with the pragmatic meaning of the original sentence. Haßler (Reference Haßler, Güldemann and von Roncador2002, p. 152) mentions the following reasons contributing to discrepancies between the reporter’s voice and that of the original speaker:
-
1. The information is generally known or part of some tradition.
-
2. The speaker has indirectly learned the fact he communicates from a third person or by hearsay.
-
3. The content of the message has been deduced.
-
4. The content of the message is the result of reflection.
The interactions provided in Extracts 5–8 indicate that the distorted reported speech,Footnote 6 that is, a speaker’s real utterance is substituted by a misled, misrepresented or exaggerated version of the utterance. Against this backdrop, both pragmatic and semantic ambiguity can create this distortion, giving the reporter the freedom to report as desired. This distorted reported speech was observed frequently among Iranian EFL learners. In Extracts 5–8, the reporter has subjectively altered the original speaker’s intention and re-states the original speaker’s utterances by changing the illocutionary meaning of the sentence. The main characteristic of those polyphonic constructs is the reformation of the semantic and syntactic features of the original speech by the reporter, and the regeneration of the original speaker’s utterance, both of which are based on the presumptions facilitated through the sociocognitive context (Morady Moghaddam, Reference Morady Moghaddam2019). In these four cases of polyphonic utterances, the original speaker is available when the report is made and refutes the reporter’s interpretation of the sentence. The distorted speech reveals that the original speaker may not be the main person responsible for what is said during the report. The distorted report suggests that both the original speaker and the reporter can be responsible for the illocutionary force of the report.
4.3. Polyphony and quasi-quotations among Iranian EFL learners
In spoken discourse, the difference between direct and indirect reports, which are generally separated in writing by the ‘quotation mark’, is not clear-cut. It means that some forms are more explicit in indicating whose voice is represented and other forms allow different interpretations with respect to whom the intended speaker is. But the choice between those forms is a syntactic one. There might be no specific marker to reveal indirect reports, and the reporter rarely, if ever, explicitly announces the introduction of an indirect report. However, Iranian EFL learners resorted to several ‘popping out’ marks in their indirect reports that signify that the polyphonic utterance may not be a fair paraphrase of the original speech. In other words, the ‘popping out’ marks underlined in extracts below mitigate the level of accuracy of the reports, leading the hearer not to consider the report as a verbatim one; according to Itakura (Reference Itakura, Capone, Garcia-Carpintero and Falzone2018, p. 328), ‘direct reports are not supposed to be used with hedging expressions’. The following extracts from the data point out how complicated the reported speech can be in L2 discourse:
Extract 9.
R: You say that, in any case, the automobile industry is sensitive.
Extract 10.
R: You say we should move towards a competitive market?
Extract 11.
R: Yesterday, the venerable Minister of Education said that, now, maybe there is no need to cancel the national entrance exam.
Extract 12.
R: Now you say that the body of the Ministry of Commerce, Iran Xodro, blah, blah, blah.
Extract 13.
R: In other words, he says these resources are not achievable.
Extract 14.
R: That is, you say there is no consistency in decision-making?
It is possible to see traces of popping-out marks and expressions in Extracts 9–14, which are perfect examples that indicate how syntax can help L2 learners to distinguish between the cooperative polyphony and polyphony beyond the original speaker’s perspective (that the report is or is not strictly quoted). The inclusion of hedges and paraphrasing marks may highlight the reporter’s uncertainty in using source information (unclear language use, according to Fløttum, Reference Fløttum2010). In Extract 9, the reporter uses ‘in any case’ as a discourse marker to imply either a slight change of the original information or to indicate that some details are ignored in the report. The complication is that it is not possible to vividly assign ‘in any case’ to the original speaker (since it is included in the ‘that’ clause of the report), while it can be regarded as a quasi-quotation (a blurred indirect report). In Extract 10, the rising intonation (the question mark) indicates an uncertain report, which indirectly manifests that the polyphony is not based on accurate information. The word ‘now’ in Extract 11, which is included in the ‘that-clause’ of the report, is another popping remark referring to unclear language use. The literal meaning of the word ‘now’ has nothing to do with the context of the reported speech in Extract 11, yet ‘now’ here is a gap filler, which suggests that the L2 reporter is trying to buy himself time to retrieve proper information. In Extract 12, using ‘blah’ three times in succession implies that the reporter has paraphrased the original information or is uninterested in quoting certain parts of the source message. As noticed by Saka (Reference Saka, Saka and Johnson2017, p. 47), ‘blah’ ‘connote[s] disinterest on the part of the reporter: material is omitted not just because it is irrelevant but, especially in the case of blah, because it is tedious, and the original source is portrayed as a blatherer’. In Extracts 13—14, the reporters employ ‘in other words’ and ‘that is’, respectively, to indicate that the polyphonic utterances are not a strict paraphrase of the original speech (there are some doubts on the part of the L2 reporter).
It can be concluded that Iranian EFL learners’ reported speech allows for some popping-out expressions (quasi-quotations) that help them to communicate that they are not strictly quoting the original speaker. Saka (Reference Saka, Saka and Johnson2017, p. 36) states that ‘[t]he English language possesses a variety of quasi-quotational devices: expressions that, when appearing inside quotation marks, conventionally trigger a temporary “popping out” of the quotation, alerting the audience that the expression is not being strictly quoted’. In this case, polyphonic utterances in Extracts 9–14 resemble what Saka introduced as the ‘quasi-quotation’. Quasi-quotation may occur for several reasons (Saka, Reference Saka, Saka and Johnson2017, p. 45):
… to harmonize the syntax of the source with that of the matrix report (especially in mixed quotation); to abbreviate the source’s message for the sake of brevity; to elide distracting bits of the source message for the sake of foregrounding others; to clarify the source; to interpolate editorial comments about the source; to save someone from embarrassment; to keep from getting in trouble; and to communicate a partial message when in ignorance of the whole. (due to obscurity in the original source, forgetfulness on the reporter’s part, or other failure)
However, popping-out marks are not linguistic devices that Iranian EFL learners frequently use in their indirect reports. Sometimes it is not feasible to identify the changes made by the reporter. Therefore, ignoring the inclusion of clues and cues is a natural phenomenon in polyphonic utterances. This can define why indirect reports are cumbersome to be identified as a polyphonic construct, for the reporter can delete all clues and cues (Capone, Reference Capone2016) and paraphrase the original sentence without taking into account the source message. This is exactly the reason for making a distinction between reporting as a pragmatic act (in which the polyphony of utterances can remain implicit) and marking reports in the language structure, that is, making the polyphony apparent in the ‘stable conventionalised patterns of the language itself’, in Voloshinov’s phrase. In compliance with McCullagh (Reference McCullagh, Saka and Johnson2017, p. 10):
One can hear some words uttered but not know what language they are in; or who is uttering them; or where the speaker is. In such a case one does not know enough about the utterance to be justified in selecting any particular lexicon (the first case) or giving any particular values to the contextual parameters. (the other cases)
This study’s findings indicated that the reporting speaker does not always report the original speaker’s utterance cooperatively in polyphonic utterances such as indirect reports. The data also showed that Iranian EFL learners included their own interpretation in their reports, so as to convey an implicature or to be strategically ambiguous, according to Fløttum (Reference Fløttum2010). Different cases of polyphony (the first and the second type of DVD) in Iranian EFL learners unequivocally revealed that L2 learners may change the intended meaning of the report based on their interpretation and evaluation of the original message, hence, creating opposing views. In the cases of distorted reported speech, the interpretation of the original utterance is illicit in such a way that it loses its resemblance with the original utterance and attaches false commitments to the original speaker (cf. Earis & Cormier, Reference Earis and Cormier2013). Therefore, in line with DVD, this study illustrated that Iranian EFL learners can (consciously or unconsciously) obfuscate their voices in order to achieve a particular communication goal. In other words, the sociopragmatic variation observed in the L2 reported speech was the result of learners’ context-specific comprehension of the original speech and a follow-up projection of their own interpretations into the indirect reports.
4.4. The distinction between the DVDs in L2 reported speech
In the case of the first type of DVD, Iranian EFL learners conform to Capone’s Paraphrase Principle. As reported by Capone (Reference Capone2010, p. 382), ‘[t]he that-clause embedded in the verb “say” is a paraphrase of what Y said if it meets the following constraint: Should Y hear what X said Y had said, Y would not take issue with it, but would approve of it as a fair paraphrase of the original utterance’. The first type of DVD manifested that L2 learners reduced the cognitive load of the original speech by maintaining the underlying meaning and transforming the original speaker’s utterance in line with the original speaker’s pragmatic force (hence, Capone’s Paraphrase Principle is supported). On the other hand, Extracts 5–8 illustrate that the second type of DVD signifies a contradiction with the illocutionary force of the original speaker’s utterance, which does not support Capone’s Paraphrase Principle as well as the idea of perfect samesaying in polyphonic utterances. In the case of the second type of DVD, Capone’s (Reference Capone2010) Paraphrase Principle is not observed in this study’s L2 data. In type two, polyphony in L2 reported speech is very close to the phenomena described by Fløttum (Reference Fløttum2010, p. 998) regarding unclear language use where ‘the sources of the different points of view that are introduced are not made explicit’.
Few Persian studies on indirect reports exist, with the literature review revealing only two relevant studies in this regard (Capone & Nodoushan, Reference Capone and Nodoushan2014; Nodoushan, Reference Nodoushan2018). Unlike previous studies on reported speech in the Iranian context (Nodoushan, Reference Nodoushan2018), this study illustrated that Iranian L2 learners may adapt the verbal tense in indirect speech. The findings of this study do not support the statements made by Capone and Nodoushan (Reference Capone and Nodoushan2014) and Nodoushan (Reference Nodoushan2018) regarding the application of Paraphrase Principle. The distorted reported speech observed among Iranian EFL learners indicates that L2 learners can change the pragmatic features of the original utterance (the second type of DVD) during the report to the extent that the report breaks its connection with the original speaker’s perspective. Having talked about several important features of L2 reported speech, Table 2 offers a summary of the data regarding the main characteristics of polyphony with regard to the first and the second type of DVD:
Table 2. Features of polyphony observed among Iranian EFL learners
5. Conclusion
Indirect reporting among Iranian EFL learners, as cases of polyphony, can be considered a joint activity (both cooperative and opposing) that is justified and accounted for based on both pragmalinguistics and sociopragmatics (Marmaridou, Reference Marmaridou, Bublitz and Norrick2011). Although indirect reporting is a pragmalinguistic approach (because it deals with illocutionary meaning), it is also a sociopragmatic paradigm since it takes into consideration ‘assessment of participants’ social distance, the language community’s social rules and appropriateness norms, discourse practices, and accepted behaviours’ (Marmaridou, Reference Marmaridou, Bublitz and Norrick2011, p. 77). We began this article by discussing some theoretical underpinnings of polyphony and language game and mentioned that changes/paraphrases and samesaying are in close proximity to the idea of polyphony. The L2 data revealed that polyphony among Iranian EFL learners consists both of cooperative (the first type of DVD) and opposing (the second type of DVD) events, that is, issues such as changes, samesaying and accountability are differently dealt with and accounted for in each type. Moreover, the data demonstrated, by referring to distorted reported speech, that samesaying may not be a hundred per cent perfect among Iranian EFL learners when the reporter changes the content of the original speaker as a way to convey a particular impression on the hearer or to create strategic ambiguity.
The discourse practices and interlanguage pragmatics also take on special meaning in the L2 classroom, as the question arises of whether language learners take on the sociopragmatic rules of their native language, which may or may not differ from the rules of others who have the target language as their native language. Blattner and Fiori (Reference Blattner and Fiori2011) make the case that educators focus on several linguistic areas such as semantics and syntax along with building skills in the different modalities of listening comprehension, speaking, reading and writing, and have only limited time, which results in the common omission of teaching sociopragmatic elements of learning. It was mentioned that the responsibility of the content conveyed through indirect reports should be analysed based on the notion of polyphony. Regarding the first type of DVD, polyphony shows that the responsibility of the report is largely on the shoulders of the original speaker since the reporter has not changed the pragmatic force of the original utterance (hence the report is in accordance with the original speaker’s perspective). Regarding the second type of DVD, the responsibility is shared between the reporter and the original speaker. And finally, we refuted the generalizability of Capone’s (Reference Capone2010) Paraphrase Principle to polyphonic utterances in L2 discourse and mentioned that the Principle is only applicable to the first type of DVD. When DVD conveys opposing views, the Principle falls short of accounting for cases of unclear language use (Fløttum, Reference Fløttum2010). Additionally, we counterargued the applicability of the Paraphrase Principle to Iranian EFL learners by discussing cases of the distorted reports observed in L2 classroom discourse. The Paraphrase Principle was not observed when the reporter manipulated the illocutionary force of the utterance beyond the original speaker’s perspective.
Overall, Iranian EFL learners’ production of reported speech contributes significantly to the knowledge of polyphony and language use by revealing that polyphonic utterances may deviate from the perspective of the original speaker. Just as this has been shown true when speakers communicate in their native language, it appears that the same holds true in advanced speakers of an L2. This work’s authentic data and relevant discussions can be considered as a guideline for other researchers interested in studying L2 reported speech, to see whether a universal pattern exists with regard to L2 reported speech. More research is needed to study polyphonic utterances as conceptualised via reported speech, particularly comparing L1 and L2 speakers in local contexts.
Data availability statement
The dataset generated during this study is not publicly available due to legal restrictions made by the participating institutes of this study but is available from the corresponding author upon a reasonable request.
Competing interests
The authors declare none.
Funding statement
This research received no specific grant from any funding agency.
